Network Communications and Protocols

chickpeasulotrichousΔίκτυα και Επικοινωνίες

27 Οκτ 2013 (πριν από 3 χρόνια και 9 μήνες)

74 εμφανίσεις

Network Communications
and Protocols
Chapter 6
2
Learning Objectives
 Understand function and structure of packets in
network, and analyze and understand these
packets
 Understand function of protocols in network
 Discuss layered architecture of protocols,
and describe common protocols and their
implementation
 Understand channel access methods
3
Function of Packets in Network
Communications
 Networks reformat data into smaller, more
manageable pieces called packets or frames
 Advantages of splitting data include:
 More efficient transmission, since large units of data
saturate network, as seen in Figure 6-1
 More computers able to use network
 Faster transmissions since only packets
containing errors need to be retransmitted
4
Large Blocks of Data Sent by One
Computer Tie Up Network
5
Packet Structure
 Three basic parts of packet, as seen in
Figure 6-2:
 Header – contains source and destination
address along with clocking information to
synchronize transmission
 Data –payload or actual data, can vary from
512 bytes to 16 kilobytes
 Trailer – information to verify packet’s contents, such
as Cyclic Redundancy Check (CRC)
6
Typical Packet Structure
7
Packet Creation
 From sender, data moves down layers of
OSI model
 Each layer adds header or trailer information
 Data travels up layers at receiver
 Each layer removes header or trailer information
placed by corresponding sender layer
 See Figure 6-3
8
Header/Trailer Information Added or
Removed
9
Packet Creation
 Outgoing data stream enters OSI model as
complete message
 Remains as data at Layers 5-7
 Lower-layers split data
 Transport Layer 4 splits it into segments
 Network Layer 3 splits segments into packets
 Data Link Layer 2 puts packets into frames
 Physical Layer 1 transmits packets as bits
10
Understanding Packets
 Three kinds of packets:
 Unicast packet - addressed to only one computer
 Broadcast packet – created for all computers
on network
 Multicast packet – created for any computers
on network that “listen” to shared network
address
11
Protocols
 Rules and procedures for communicating
 To communicate, computers must agree
on protocols
 Many kinds of protocols:
 Connectionless
 Connection-oriented
 Routable
 Nonroutable
12
The Function of Protocols
 Each protocol has different purpose and function
 Protocols may work at one or more layers
 More sophisticated protocols operate at higher
layers of OSI model
 Protocol stack or protocol suite is set of
protocols that work cooperatively
 Most common protocol stacks are TCP/IP used
by the Internet and IPX/SPX used by Novell
NetWare
13
Connectionless Versus Connection-
Oriented Protocols
 Two methods for delivering data across network:
 Connectionless – no verification that datagrams
were delivered; fast protocols with little overhead
 Connection-oriented – more reliable and slower
protocols that include verification that data was
delivered; packets resent if errors occur
14
Routable Versus Nonroutable Protocols
 Network Layer 3 moves data across multiple
networks using routers
 Routable – protocols that function at Network
layer, such as TCP/IP or IPX/SPX, essential for
large-scale networks or enterprise networks
 Nonroutable – protocols that do not include
Network layer routing capabilities, such as
NetBEUI, work well in small network
 Consider current size and future expansion
possibilities when choosing protocol suite
15
Protocols in a Layered Architecture
 Most protocols can be positioned and explained
in terms of layers of OSI model
 Protocol stacks may have different protocols for
each player
 See Figure 6-4 for review of functions of each
layer of OSI model
 See Figure 6-5 for three major protocol types
 Application protocolsat Layers 5-7
 Transport protocols at Layer 4
 Network protocolsat Layers 1-3
16
Functions of OSI Model Layers
17
Three Main Protocol Types
18
Network Protocols
 Provide addressing and routing information, error
checking, and retransmission requests
 Services provided by network protocols are called link
services
 Popular network protocols include:
 Internet Protocol (IP)
 Internetwork Packet Exchange (IPX) and NWLink
 NetBEUI
 Delivery Datagram Protocol (DDP)
 Data Link Control (DLC)
19
Transport Protocols
 Handle data delivery between computers
 May be connectionless or connection-oriented
 Transport protocols include:
 Transmission Control Protocol (TCP)
 Sequenced Packet Exchange (SPX) and NWLink
 AppleTalk Transaction Protocol (ATP) and
Name Binding Protocol (NBP)
 NetBIOS/NetBEUI
20
Application Protocols
 Operate at upper layers of OSI model to provide
application-to-application service
 Some common application protocols are:
 Simple Mail Transport Protocol (SMTP)
 File Transfer Protocol (FTP)
 Simple Network Management Protocol (SNMP)
 NetWare Core Protocol (NCP)
 AppleTalk File Protocol (AFP)
21
Common Protocol Suites
 TCP/IP
 NWLink (IPX/SPX)
 NetBIOS/NetBEUI
 AppleTalk
 DLC
 XNS
 DECNet
 X.25
Combination of protocols that work
cooperatively to accomplish network
communications
Some of the most common protocol suites
are:
22
Transmission Control Protocol/ Internet
Protocol (TCP/IP
 Called the Internet Protocol (IP)
 Most commonly used protocol suite for networking
 TP/IP used by US Department of Defense’s Advanced
Research Projects Agency (ARPA)
 Excellent scalability and superior functionality
 Able to connect different types of computers and
networks
 Default protocol for Novell NetWare, Windows 2000/XP,
and Windows NT
 See Figure 6-6 for relationship to OSI model
23
TCP/IP Compared to OSI Model
24
TCP/IP
 Includes highly compartmentalized and
specialized protocols, including:
 Internet Protocol (IP) – Connectionless Network
layer protocol that provides source and destination
routing; fast, but unreliable
 Internet Control Message Protocol (ICMP) –
Network layer protocol that sends control messages;
PING uses ICMP
 Address Resolution Protocol (ARP) – Network
layer protocol that associates logical (IP) address to
physical (MAC) address
25
More TCP/IP Protocols
 Transmission Control Protocol (TCP) – primary
Internet transport protocol; connection-oriented; provides
reliable delivery; fragments and reassembles messages
 User Datagram Protocol (UDP) - connectionless
Transport layer protocol; fast, unreliable
 Domain Name System (DNS) – Session layer
name-to-address resolution protocol
 File Transfer Protocol (FTP) – performs file transfer,
works at Session, Presentation, and Application layers
26
More TCP/IP Protocols
 Telnet – remote terminal emulation protocol; operates at
three upper layers; provides connectivity through
dissimilar systems
 Simple Mail Transport Protocol (SMTP) –
operates at three upper layers to provide messaging;
allows e-mail to travel on Internet
 Routing Information Protocol (RIP) – Network layer
distance-vector protocol used for routing;
not suitable for large networks
 Open Shortest Path First (OSPF) – link-state routing
protocol; uses variety of factors to
determine best path
27
IP Addressing
 Logical addresses, 32-bits or 4 bytes long
 Four octets separated by periods, each with
decimal value from 0-255
 First part of address identifies network
 Second part of address identifies host or
individual computer
 IP addresses broken into classes
 Number of IP address registries under control of
Internet Assigned Numbers Authority (IANA)
28
IP Address Classes
 Three classes of IP addresses for normal
networking:
 Class A– addresses between 1-126; first octet
identifies network and last three identify host;
over 16 million hosts per network
 Class B – addresses between 128-191; first
two octets identify network and last two identify host;
over 65,000 hosts per network
 Class C – addresses between 192-223; first
three octets identify network and last one
identifies host; limited to 254 hosts per network
29
IP Address Classes
 Two classes of IP addresses have special
purposes:
 Class D – addresses range from 224-239;
reserved for multicasting; used for videoconferencing
and streaming media
 Class E – addresses range from 240-255;
reserved for experimental use
30
Special Service IP Addresses
 Some addresses used for special services:
 IP addresses beginning with 127 are loopback
addresses; also called localhost
 Reserved addresses for private networks
include:
 Class A addresses beginning with 10
 Class B addresses from 172.16 to 172.31
 Class C addresses from 192.168.0 to 192.168.255
31
IPv6
 Current four byte version is IPv4
 Now reaching limit of 4-byte addresses
 IETF working on new implementation of TCP/IP,
designated IPv6
 Uses 16 byte addresses
 Retains backward compatibility with IPv4
4-byte addresses
 Will provide limitless supply of addresses
32
Classless Inter-Domain Routing (CIDR)
 Internet uses CIDR
 Demarcation between network and host not
always based on octet boundaries
 May be based on specific number of bits
from beginning of address
 Called subnetting, the process involves
“stealing” bits from host portion of address
for use in network address
 Provides fewer hosts on each networks but
more networks overall
33
Subnet Masks
 Part of IP address identifies network and part
identifies host
 IP uses subnet mask to determine what part
of address identifies network and what part
identifies host
 Network section identified by binary 1
 Host section identified by binary 0
34
Subnet Masks
 Each class of addresses has default subnet
mask
 Class A default subnet mask is 255.0.0.0
 Class B default subnet mask is 255.255.0.0
 Class C default subnet mask is 255.255.255.0
 All devices on single physical network or
network segment must share same network
address and use same subnet mask
35
Some Simple Binary Arithmetic
 Four kinds of binary calculations:
 Converting between binary and decimal
 Converting between decimal and binary
 Understanding how setting high-order bits to value of
1 in 8-bit binary numbers corresponds
to specific decimal numbers
 Recognizing decimal values for numbers that
correspond to low-order bits when they’re set
to value of 1
 Keep in mind that any number raised to
zero power equals one
36
Converting and Understanding High- and
Low- Bit Patterns
 Converting Decimal to Binary
 Divide number by 2 and write down remainder which
must be 1 or 0
 Converting Binary to Decimal
 Use exponential notation
 High-Order Bit Patterns
 See Table 6-1
 Low-Order Bit Patterns
 See Table 6-2
37
High-Order Bit Patterns
38
Low-Order Bit Patterns
39
Calculating a Subnet Mask
 Follow these steps to build subnet mask:
 Decide how many subnets you need
 Add two to number of subnets needed (one for
network address and other for broadcast address).
Then jump to next highest power of 2
 Reserve bits from top of host portion of address down
 Be sure enough host addresses to be usable are
left over
 Use formula 2
b
– 2 to calculate number of usable
subnets, where b is number of bits in subnet mask
40
Calculating Supernets
 Supernetting “steals” bits from network portion
of IP address
 Supernets permit multiple IP network addresses
to be combined and function as a single logical
network
 Permit more hosts to be assigned on supernet
 Improves network access efficiency
41
Network Address Translation (NAT)
 Allows organization to use private IP addresses
while connected to the Internet
 Performed by network device such as router that
connects to Internet
 See Figure 6-7 for example of NAT
42
Network Address Translation (NAT)
43
Dynamic Host Configuration Protocol
(DHCP)
 DHCP server receives block of available
IP addresses and their subnet masks
 When computer needs address, DHCP server
selects one from pool of available addresses
 Address is “leased” to computer for designated length
and may be renewed
 Can move computers with ease; no need to
reconfigure IP addresses
 Some systems, such as Web servers, must have
static IP address
44
NetBIOS and NetBEUI
 Consortium of Microsoft, 3Com, and IBM
developed lower-level protocol NetBEUI in mid-
1980s
 NetBIOS Extended User Interface
 Spans Layers 2, 3, and 4 of OSI model
 Both designed for small- to medium-sized
networks, from 2-250 computers
45
NetBIOS and NetBEUI
 Figure 6-8 shows Microsoft protocol suite and its
relationship to OSI model
 Defines four components above Data Link layer
 Runs on any network card or physical medium
 Redirector interprets requests and determines whether
they are local or remote
 If remote, passes request to Server Message Block
(SMB)
 SMB passes information between networked
computers
46
Microsoft Protocol Suite Compared to
OSI Model
47
NetBIOS and NetBEUI
 NetBEUI works at Transport layer to manage
communications between two computers
 Nonroutable protocol; skips Network layer
 NetBEUI packet does not contain source or
destination network information
48
NetBIOS and NetBEUI
 NetBIOS operates at Session layer to provide
peer-to-peer network application support
 Unique 15-character name identifies each computer
in NetBIOS network
 NetBIOS broadcast advertises computer’s name
 Connection-oriented protocol, but can also use
connectionless communications
 Nonroutable protocol, but can be routed when using
routable protocol for transport
49
NetBIOS and NetBEUI
 NetBEUI is small, fast, nonroutable
Transport and Data Link protocol
 All Windows versions include it
 Ideal for DOS based computers
 Good for slow serial links
 Limited to small networks
 Server Message Block operates at
Presentation layer
 Used to communicate between redirector
and server software
50
IPX/SPX
 Original protocol suite designed for Novell’s
NetWare network operating system
 Still supported with NetWare 6.0, but TCP/IP
is now primary protocol
 NWLink is Microsoft’s implementation of
IPX/SPX protocol suite
 Figure 6-9 shows protocols in NWLink and
corresponding OSI layers
 Must consider which Ethernet frame type with
NWLink
51
NWLink Compared to
OSI Model
52
IPX/SPX
 Open Data-link Interface (ODI) lets single
network driver support multiple protocols
through single NIC
 Internetwork Packet Exchange (IPX) is
Transport and Network layer protocol
 Handles addressing and routing
 Connectionless protocol
 Provides fast, but unreliable, services
53
IPX/SPX
 Other protocols in the IPX/SPX suite include:
 IPX Routing Information Protocol (IPX RIP) –
distance-vector protocol; uses ticks to determine best
path; exchanges information about network
addresses and topology
 Sequenced Packet Exchange (SPX) – provides
connection-oriented service; more reliable
 NetWare Core Protocol (NCP) – works at Transport
and upper layers to provide range of client -server
functions
54
IPX/SPX
 Other protocols in IPX/SPX suite include:
 Service Advertising Protocol (SAP) – used by file
and print servers to advertise services
 Service Lookup Protocol (SLP) – new IP-based
NetWare protocol used with Novell Directory
Services; used when clients want to look up services
on IP-only network
55
AppleTalk
 Defines physical transport in Apple
Macintosh networks
 Divides computers in zones
 AppleTalk Phase II allows connectivity outside
Macintosh world
56
Xerox Network Systems (XNS)
 Designed for Ethernet networks
 Basis for Novell’s IPX/SPX
 Rarely used in today’s networks
57
DECNet
 Used with Digital Network Architecture
 Proprietary protocol
 Complete routable suite
 Phase IV closely resembles OSI model
58
X.25
 Set of wide-area protocols
 Designed to connect remote terminals to
mainframes
 Used in packet-switching networks
 Still widely used in Europe
59
Implementing and Removing Protocols
 Easy to add or remove protocols
 TCP/IP loads automatically when most operating
systems are installed
 In Windows 2000/XP, use Network and
Dial-up Connections control panel
 See Figure 6-10
60
Network and Dial-up Connections
61
Putting Data on the Cable: Access
Methods
 Consider several factors
 How computers put data on the cable
 How computers ensure data reaches destination
undamaged
62
Function of Access Methods
 Rules specify when computers can access cable
or data channel
 Channel access methods assure data reaches
its destination
 Prevents two or more computers from sending
messages that may collide on cable
 Allows only one computer at a time to send data
63
Major Access Methods
 Channel access is handled at Media Access
Control (MAC) sublayer of Data Link layer
 Five major access methods
 Contention
 Token passing
 Demand priority
 Polling
 Switching
64
Contention
 In early networks, contention method allowed computers
to send data whenever they had
data to send, resulting in frequent collisions and
retransmissions
 Figure 6-11 shows data collision
 Two carrier access methods were developed for
contention-based networks
 Carrier Sense Multiple Access with Collision
Detection (CSMA/CD)
 Carrier Sense Multiple Access with Collision
Avoidance (CSMA/CA)
65
Data Collision
66
CSMA/CD
 Popular access method used by Ethernet
Prevents collisions by listening to channel
If no data on line, may send message
If collision occurs, stations wait random period
of time before resending data
See Figure 6-12
67
CSMA/CD
68
CSMA/CD
 Limitations and disadvantages of CSMA/CD
Not effective at distances over 2500 meters
More computers on network likely to cause
more collisions
Computers have unequal access to media
Computer with large amount of data can
monopolize channel
69
CSMA/CA
 Uses collision avoidance, rather than detection,
to avoid collisions
 When computer senses channel is free, it signals its
intent to transmit data
 Used with Apple’s LocalTalk
 Advantages and disadvantages
 More reliable than CSMA/CD at avoiding collisions
 “Intent to transmit” packets add overhead and reduce
network speed
70
Token Passing
 Token passes sequentially from one computer to next
 Only computer with token can send data, as seen in
Figure 6-13
 Advantages and disadvantages
 Prevents collisions
 Provides all computers equal access to media
 Computer must wait for token to transmit, even
if no other computer wants to transmit
 Complicated process requires more expensive
equipment
71
Communication in a
Token-Passing Network
72
Demand Priority
 Used only by 100VG- AnyLAN 100 Mbps Ethernet
standard (IEEE 802.12)
 Runs on star bus topology, as seen in Figure 6-14
 Intelligent hubs control access to network
 Computer sends hub demand signal when it
wants to transmit
 Advantages and disadvantages
 Allows certain computers to have higher priorities
 Eliminates extraneous traffic by not broadcasting
packets but sending them to each computer
 Price is major disadvantage
73
Demand Priority Uses
Star Bus Topology
74
Polling
 One of oldest access methods
 Central controller, called primary device, asks
each computer or secondary device if it has data
to send, as seen in Figure 6-15
 Advantages and disadvantages
 Allows all computers equal access to channel
 Can grant priority for some computers
 Does not make efficient use of media
 If primary device fails, network fails
75
Primary Device Controls Polling
76
Switching
 Switch interconnects individual nodes and controls
access to media
 Switching usually avoids contention and allows
connections to use entire bandwidth
 Other advantages include
 Fairer than contention-based technology
 Permits multiple simultaneous conversations
 Supports centralized management
 Disadvantage include
 Higher cost
 Failure of switch brings down network
77
Choosing an Access Method
 Network topology is biggest factor in choosing
access method
 Ring topology usually uses token-passing
 Switching can emulate all common topologies
 See Tables 6-3 through 6-7 for summaries of the
five access methods
78
Contention Access Method
79
Token-Passing Access Method
80
Demand Priority
Access Method
81
Polling Access Method
82
Switching Access Method
83
Chapter Summary
 Data stream on a network is divided into packets
to provide more reliable data delivery and ease
network traffic
 If errors occur during transmission, only packets
with errors will be re-sent
 As data travels through layers of OSI model,
each layer adds its own header or trailer
information to packet
84
Chapter Summary
 As receiving computer processes packet, each
layer strips its header or trailer information
and properly re-sequences segmented message
so that packet is in original form
 Many protocols are available for network
communications
 Each protocol has strengths and weaknesses
 A suite, or stack, of protocols allows a
number of protocols to work cooperatively
85
Chapter Summary
 Major protocol suites are TCP/IP, IPX/SPX, and
NetBEUI
 Each suite contains many smaller protocols,
each of which has its own network function
 IP addressing involves several concepts,
including address classes, subnetting,
supernetting, and subnet masks
86
Chapter Summary
 Current method for Internet addressing is called
CIDR, which uses all available addresses more
efficiently
 Other IP addressing concepts include:
DHCP, a method for automatic assignments
and management of IP addresses
NAT, which allows companies using private IP
addresses to access the Internet and use
public IP addresses more efficiently
87
Chapter Summary
 When a computer is ready to send data, it must
be assured that data will reach destination
 Perfect environment does not exist where all
computers can have dedicated channel over
which to send information
 Rules have been established to ensure that all
computers have time on the channel
 Token passing and polling guaranteed time
for each computer to send its data
88
Chapter Summary
 Demand priority allows computer to send
data after it notifies controlling hub
 In contention channel access methods,
computers vie for network time
 They listen to network to determine whether another
computer is sending data
 If not, they send their data (CSMA/CD) or broadcast
their intention to send data (CSM/CA)
 Switching can emulate all other access methods
and offers greatest total available bandwidth